Critique argues GPUI and Egui state models are unsuitable for complex desktop apps

Concerns with Rust GUI Frameworks

A recent critique highlights concerns regarding the suitability of Rust GUI frameworks like GPUI and Egui for developing complex desktop applications. While these frameworks are gaining traction, with Discord rewriting its app using GPUI and many LLM-clones using Egui, the author argues their underlying state models present significant limitations for larger projects.

Scalability Issues with State Models

The core argument is that both GPUI's and Egui's state models are 'dead ends' for complex applications. The author contends that every UI architecture eventually needs to track what changed and who depends on it for performance scalability. Frameworks that do not inherently support this force applications to manage these aspects manually, which negates the initial appeal of their API semantics.

The critique suggests that once precise change tracking is required, components that run once, utilizing signals and reactive holes, offer a simpler API for achieving this.

GPUI's Hybrid Approach and Limitations

GPUI is described as a 'hybrid immediate and retained mode' framework that retains state in entities and defines the UI in render functions. While it uses ordinary Rust without special wrappers, state changes necessitate an explicit rerun. The author argues that while such approaches appear simple initially due to no state synchronization and plain Rust render functions, they do not scale for complex applications, and attempts to fix these issues undermine their original simplicity.

Implications for Developers

This analysis implies that developers considering GPUI or Egui for new desktop applications, especially those with complex state management requirements, should carefully evaluate these frameworks' long-term scalability. While suitable for debug interfaces or simpler applications, their architectural choices may lead to performance and maintenance challenges in more intricate projects.